This project will capitalize on novel microelectrode recording technology and the unique advantages of the transgenic mouse system to study persistent neural activity. Head Direction (HD) cells, the circuit studied, integrate ideothetic cues such as angular head velocity to provide a neural firing pattern that represents head direction in the lab frame of reference. Because the activity remains when sensory inputs are restricted, the activity has the characteristics of a short term memory. Neural network models describe this system, using a framework pioneered by Hopfield, as a continuous attractor driven by net positive feedback based on recurrent excitation. The research objective is to measure the network and cellular mechanisms underlying persistent neural activity in the HD system, relating these properties to recurrent network models. Now a novel recording method developed here can relate for the first time the firing properties of multiple single HD cells in the same small nucleus of awake and behaving mice, where each of up to five individually adjustable electrodes can simultaneously record an isolated single neural unit. This device allows measurement of the persistence time of the network and the timing relationship between similarly tuned HD cells. Developing this mouse recording system allows us to use genetics to overexpress or delete important ion channels and receptors for this system such as the N-methyI-D-aspartate receptor. This work will combine in vivo electrophysiology, transgenic technologies, pharmacology, viral infusions and other methods to manipulate and measure the system to reveal its properties. The work will be a collaboration between the labs of Drs. Tank and Tsien at Princeton University and consultant Dr. J. S. Taube. This project has broad significance for neuroscience for two reasons. Firstly, the electrode method development here is the first to allow measurement of multiple single neurons in any small nucleus of a mouse, and should find applicability by other laboratories studying other neural systems. Recording methods for mice will become even more crucial as transgenic mouse technologies advance. Secondly, the continuous attractor model studied here for HD cells has also been proposed for a variety of other systems, so the cellular and network mechanisms elucidated here could very well apply to short-term memory in general.